DESCRIPTION: This application is from Memorial Sloan-Kettering (MSK). Previous 31-P NMR spectroscopic studies of murine tumors has suggested that it may be feasible to predict tumor responses from spectral changes. Preliminary clinical studies from many institutions, studying a range of tumors, using different scanners, with a plethora of varied techniques, have suggested that although the spectral changes in patients and murine tumors appear to be quite different, nevertheless, it may be possible to predict clinical tumor response on the basis of spectral changes. In some studies it has been suggested that spectra obtained pretreatment may be indicative of tumor response. The applicants are therefore proposing a multi institutional cooperative type study to answer the questions of whether NMR spectroscopy can 1) a priori predict tumor response (from the pretreatment spectrum) and 2) be an early predictor of tumor response by monitoring spectral changes during the first cycle of chemotherapy prior to tumor shrinkage. The applicants propose to use inform techniques (coils, data analysis, acquisition techniques, 1H decoupling, volume localization etc) so that the spectra are comparable from one institution to the other. The applicants propose to study four different tumors (extremity sarcomas, advanced breast cancer, non Hodgkin's lymphoma, and squamous cell cancer of the head and neck). These tumor sites have the common factor of being superficial and therefore it is expected that localization will be easier (and more accurate) and signal to noise better. These two factors are critical if one is to accurately measure tumor metabolite concentrations. The hypotheses enumerated above will be tested by correlating spectral changes with tumor response, disease free survival, survival, and other clinical markers that are being obtained on these patients. In addition to addressing these important clinical questions, this study should also provide insight into phospholipid metabolism by monitoring changes in phospholipid anabolite and catabolite. Phospholipid metabolism may be an important factor in determining tumor growth, cell cycle distribution, metastases and other important facets of tumor biology and the NMR studies may yield new insights into this area.